Downsampling

In yesterday’s post 192kHz? I explained how we wound up with the different sample rates.  A convoluted story at best.  Today I thought it might be interesting if I touched on a little known problem with pretty much all CD’s.  Downsampling.

Nearly all master digital studio recordings are made at higher sample rates and bit depths than a CD can handle, even if it’s only 48kHz.  For example, Keith Johnson at Reference records all their materials at 176.4kHz/24 bits and nearly every recording studio is making masters at 96kHz/24 or higher.  When a recording engineer is making a master recording of a group, a copy of a master analog tape etc. he is naturally going to make that master at the best possible resolution he can.  Not at the lowest needed.

So what happens when we want to go make a CD version of that master tape?  The original master is downsampled to 44.1kHz/16 bits through a sample rate converter.  Depending on what the engineer started with, downsampling clearly loses some of what was in the original recording in the process.

How much of what we recognize as inferior playback of the CD, relative to high rez media, is due to this downsampling process?

This practice, I am convinced, is one reason many CD’s sound worse than their high resolution counterparts.  If you read any of the thoughts by recording engineers on this practice, there are many that believe recording at higher rates and downsampling sounds better than keeping everything at the intended use rates.  And then there are those that prefer upsampling.  Clearly, there’s not much in the way of consensus.

In a conversation with Keith Johnson (whose ears I trust completely) he told me the single worst thing that ever happened to any of his work was the downsampling of the master tapes to CD’s.  He hated the results and I am sure this must have prompted their subsequent release of their HRx series with exact copies of the masters.  I’ll bet our friend Cookie Marenco of Blue Coast might have similar misgivings about downsampling.

The difference between the CD and HRx version played on a PWT is obvious and immediate.  But I don’t believe it has as much to do with the increase in dynamic range and frequency response afforded by the higher resolution format as it does by the lack of downsampling.

Downsampling may be one of the worst things to ever happen to the CD and one of the central reasons we judge it so harshly.

It may also partially answer why we can make a digital copy of an analog event without much in the way of loss.  We don’t have to mess with the final result of the conversion, just play it back at whatever sample and bit rate we recorded it at.

We never downsample.

Paul McGowan – PS Audio, Intl.

192kHz?

Have you ever wondered why there are so many different sample rates and why?  The compact disc samples at 44.1kHz/16 bits.  But then there’s these other rates like 48kHz, 88.2kHz, 96kHz, 176.4kHz, 192kHz.  Wow, that’s a lot and the spacing doesn’t seem to make a lot of sense.  How did we get to these numbers anyway?

The first thing to understand is there are basically only two main sample frequencies: 44.1kHz and 48kHz.  Everything else is a variant of one of those two numbers.  96kHz is a doubling of 48kHz, while 192kHz is again doubling that number.  88.2kHz and 176.4kHz are doubled numbers of the original CD standard of 44.1kHz.  But why do we have two numbers, 44.1kHz and 48kHz as the standards?  After all, these are pretty darned close to each other.  Doesn’t make much sense.  You’re right.

There’s a lot of opinions on how we got two primary sample rates so closely spaced together and here’s the best I can make out the history.  It’s an interesting story filled with a lot of twists and turns and we have space for only a few of them.

Back in the late 1970′s and early 1980′s, when digital audio began, most of the recorders were all video based.  I can remember owning a Sony PCM F1 which I used for some live audio recordings.  These rotating head video recorders ran (if memory serves correctly) at 44.1kHz (based on the highest sample rates they could handle at the time) through an adapter made by Sony and at the time were the fastest machines around.  I do recall some debate about using VHS vs. Sony’s Betamax for recording and there was a flurry of opinions and machines one could use.

Then came DAT (Digital Audio Tape also known as R-DAT), which was actually a small and convenient digital audio tape recorder by Sony (among others) that also used the rotating head video recording method to capture audio.  About the same time another variant of DAT called Soundstream digital came along for pro applications – the first use commercially was by Telarc.  Both of these machines were able to record at a slightly higher sample rate than 44.1kHz, the DAT at 48kHz and the Soundstream at 50kHz.

Simultaneously, engineering organizations and manufacturers were both trying to set standards for the new digital audio mediums.  The venerable Audio Engineering Society (AES) in concert with the European Broadcasting Union (EBU) were busy publishing standards that would allow digital audio to be sent without issue between different manufacturer’s machines.  Out of all the is work we got what we now know as the AES/EBU standard (balanced XLR) and the S/PDIF (Sony Philips Digital Interconnect Format) through a single RCA coax or optical cable.  Both formats very close to the same and both handling 44.1kHz and 48kHz plus any variant over those two frequencies.

While all this was going on, the professional recording and film studios were also getting interested in this new digital audio format for their own purposes and most of these pro applications gravitated towards the top end of the DAT sample rate possible, which was 48kHz.

So here we had a situation where Sony and Philips were pushing hard on the 44.1kHz Compact Disc standards and the pro and film markets were happy with a slightly higher sample rate of 48kHz.  It is widely rumored that the Compact Disc 44.1kHz standard evolved in order to allow a double album’s worth of music to be placed on the small silver disc.  One story even suggesting the head of Sony demanded Beethoven’s entire 9th symphony fit onto the small disc (about the same length as a double album).  Others debunk both stories as absurd since all the engineers would have to do to fit this much data with 48kHz is make the CD larger – about the size of a 45 rpm vinyl disc – and then the Sony marketing people wanting to have no relationship in size, color or look of anything vinyl making the decision.  Best I can tell it was a simple holdover from the original PCM video recorders which couldn’t handle anything higher than 44.1kHz/16 bits (the minimum required for 20Hz to 20kHz reproduction).

On the pro and film side the now common place 48kHz sample rate, a defacto “standard”, appeared on a growing number of pro-side tools and equipment and used without a lot of question.  Popular lore had it that the film industry used (and still uses as the DVD standard) 48kHz to “synch” to the standard film rate of 24fps (frames per second), but as I think about this I cannot even envision how this would matter at all – being just a convenient doubling of 24.  Synchronizing the audio to the video is essential, but the relationships between 24fps and 48kHz doesn’t seem to actually have any merit – other than that’s what everyone does.

So, in the end, we wind up with two very similar yet different primary sample rates, 44/1kHz and 48kHz.  From these two numbers come all other PCM sample rates and that, my friends, is why we have 6 types of sample rates on DACS today.  Three would be just fine, but because of the double standards of 44.1 and 48, we have 6.  Go figure.

Paul McGowan – PS Audio, Intl.

That’s a pity

In response to Thursday’s post People Like Us, reader Bob Whitlock sent me this charming note and I wanted to share it.

“I’ve always been interested in food, literally started cooking as an eight-year-old, nearly 50 years ago. At age 33, I was able to attend a 10-month culinary school in Cambridge, MA and graduated with a chef’s diploma. So, now I’m a food “audiophile” and the unintended consequence is that certain friends were quick to tell me—almost proudly—that they have very pedestrian tastes in food and, of course, would never want to serve me anything they had cooked since I would be so critical.

Ahem. Then I would ask them, “How do you like your eggs cooked?” Their reply would be specific: fried, but not browned, the yolks runny, flipped once, butter as the cooking oil, salt, very finely ground pepper. “How about hamburgers?” Oh, even more detail—seeded buns, medium rare, sweet onion, pickles, no mayo—you get the picture. And that’s when I would shake my head and tell them, “Wow, I’m not sure I would want to try to cook for someone who has such refined tastes.”

It’s ridiculous to think that someone can’t appreciate great food or great sound. They may not gravitate to either with the passion we display, but it’s within us all, just like our response to works of art or athleticism. We get it, but may not want it. And that’s a pity.”

Couldn’t have said it better myself.  Thanks Bob.


Paul McGowan – PS Audio, Intl.

Proven: CD’s sound the same as high-rez!

Incontrovertible double-blind listening tests prove that the original 16-bit/44.1-kHz CD standard yields exactly the same two-channel sound quality as the SACD and DVD-A technologies.

In the September 2007 issue of the Journal of the Audio Engineering Society (Volume 55, Number 9), two veteran audio journalists who aren’t professional engineers, E. Brad Meyer and David R. Moran, present a breakthrough paper that contradicts all previous inputs by the engineering community. They prove beyond a shadow of a doubt, with literally hundreds of double-blind listening tests at matched levels, conducted over a period of more than a year, that the two-channel analog output of a high-end SACD/DVD-A player undergoes no audible change when passed through a 16-bit/44.1-kHz A/D/A processor. That means there’s no audible difference between the original CD standard (“Red Book”) and 24-bit/192-kHz PCM or 1-bit/2.8442-MHz DSD.

Now how’s that for a good morning wake up call?  The quote I just printed from The Audio Critic (Peter Aczel) in a 2007 article.  One of our Community Forums posters was kind enough to post this for us.

So to me this is a perfect case of the wrong conclusion to an interesting test.  A test we’ve been talking about for some time as of late.  If you’ll recall, I spoke about recording a vinyl LP onto digital and how that process captured and preserved all the vinyl qualities LP’s have to offer.  In fact, using a 96kHz/24 bit A/D/A (slightly higher than Redbook) one can go between the original analog phono output and the output of the A/D/A with little audible change.  You’ll also recall that the conclusion drawn from this was not that there’s no difference between vinyl and CD, but something altogether different.  A point completely missed by the two journalists that performed the test and the Audio Critic who reported it.

To conclude that redbook CD sounds the same as High-Rez formats because you’ve compared the analog output of a High-Rez player vs. a 44.1kHz/16 bit A/D/A output of the same is not only dumb, but it misses a marvelous chance to actually say something of value.

It’s a shame they spend an entire year proving something we could have told them in a phone conversation.


Paul McGowan – PS Audio, Intl.

People like us

I certainly never have a problem when someone labels me an Audiophile although I don’t think of myself as one.  In fact, I am not sure I even know what an Audiophile is.  I am certainly a tweak, an obsessed improver, passionate about sound, an inventor, someone who cherishes time with my stereo system, a music lover, a gear lover, excited to show off my system, but an Audiophile?

We probably all know people who take a great deal of pride in proclaiming they are not Audiophiles, as if the very label itself was a poisonous one.  Yet those same people have stereo systems that would make the eyes cross of an MP3 playing pedestrian.

We probably also know people who proudly proclaim their Audiophileness to anyone who cares to listen to them (or their systems).

And there are those who eschew all the monickers and proclaim they are interested in nothing more than just the music itself.  And these same people have built great systems to reproduce this music that they are quick to point out they are only interested in.

People like us are hard to label, but I believe we all share at least two things in common: a love of music and an eagerness to share our passion of high quality reproduction.

There are billions of people in this world that share our love of music and an equal number that would appreciate a higher quality of reproduction if they heard it.

Which means there are billions of people just like us: they just don’t know it.

Paul McGowan – PS Audio, Intl.

The stories we tell ourselves

A good friend of many years is in town for a few days and we’ve been discussing high-end audio, products, the marketplace, etc.  He’s been spending time in our facilities and sees the extraordinary effort that goes into designing some of our products: the months of hard work it takes to get a product from the idea in one’s head, the napkin sketch of the circuit, the mechanical renderings of the form factor, to the final documentation needed to actually produce the product reliably and repeatably.  The level of work and detail is surprising to him.  It is surprising to me even after all these years.

When we look at a finished product rarely do we think of what it takes to successfully bring that product into the world.  I am fascinated by this because I am sure it takes about the same amount of work to bring a lower cost product into the world as it does a higher cost product.  It may be, in fact, that the best lower cost products are more difficult to birth than the fancier exotic ones – the choices that need to be made for lower cost products probably far exceed those of the price-no-object devices.

And yet when we look at a low cost “entry level” high-end audio product vs. a “cost no object” piece, we tend to shrug at the one and ooh and ahh over the other.  The time, energy, skill and inspiration to engineer both, close to the same.

The story we tell ourselves is that the entry level is a banged out cookie-cutter design while the cost-no-object is a labor of love; the vessel the designer pours his soul into.

Great stories are always fascinating.


Paul McGowan – PS Audio, Intl.

Transitions

The transition from one format to another, say vinyl to CD, CD to streaming audio, PCM to DSD, embodies the future directions music reproduction is taking.  Inevitably the first transition period has both major benefits and drawbacks as well.  We should not judge a format at the time of its transition.

The first time I ever had a chance to hear a CD was right at the transition between tried-and-true vinyl and “perfect sound forever” (a marketing phrase used by Sony – one I’ll bet they regret using – as its been used to ridicule them ever since).  This happened in the early 1980′s at the very height of the vinyl movement.  Records were king.  Everyone that listened to music had big libraries, tweaked out systems of turntables, preamplifiers, speakers that worked with the technology of the day.  Then along comes this new kid on the block boasting better everything – when actually the truth was it had the POTENTIAL for better everything.

The first CD’s and players were simply dreadful.  My fist listen was through a Magnavox player, made by Phillips.  At the time the general consensus was the Phillips players sounded better than the Sony’s – they both sounded awful (little did we know it was also the early CD’s themselves).  But despite the poor sound we could see that technically, this WAS better, it could be a lot better it just wasn’t.  So we set out to find the problems.  The first was the output stages of these early CD players.  It looked like they were added as an afterthought.  My guess is that’s accurate.  We replaced them, started the idea of a high-end modified version that was better.  Truth was it still sucked relative to the tried and true standard of the day: vinyl.

Many folks made their choices about digital audio there and then.  Never to return.  How unfortunate for them.

Technologies need time to mature.  To settle in.  Breathing room to do their work until the market’s ready to appreciate their benefits.

The benefit to taking your time to forming an opinion is a very broad playing field of enjoyment.

Paul McGowan – PS Audio, Intl.

100% anything?

I have always found it curious when someone says to me “I am 100% vinyl in my system.  No use for digital audio whatsoever.”  Or, “I never listen to vinyl, too dated, sounds bad.”  Or, “I have sold all my CD’s and gone just to SACD.”

It’s curious because these extremes many of us go towards seem limiting and unnecessary.   I know we’re an extreme lot and it takes all kinds.  I respect that and find myself in the same position time and again.

But I gently kick myself every time I fall into the trap.  For the longest time I did not have vinyl in my system.  I spent decades working on vinyl, building better and better phono preamplifiers, studying how they work, what’s the best way to deal with a phono cartridge, building magic boxes that turned moving magnets into moving coils, the list goes on.  When CD finally came of age, the pleasures and sound quality were so wonderful I just sold everything vinyl and moved on.

The first time I heard a high resolution audio bit played on the PWT I thought to myself “I can never go back”.  CD’s sounded lame to me.  High resolution was just everything.

Over time I have added back my vinyl, I listen to CD’s, I stream all manner of music to the system, I play high resolution audio.

Maybe it’s just me, but I tend to be an extreme person, swinging the pendulum back and fourth.  Over time, the pendulum swings get shorter, and my listening pleasures increase.  How much fun to grab a vinyl disc and enjoy the pleasures within, to then hop on my favorite CD, to punch up the next track on my iPad.  I just love it when I play a track for someone and they say “wow, was that high resolution?”  ”Nope, plain old CD.”

There are many mediums and if you can manage to have as many optimized systems in place to enjoy the best of what’s there, I think you’re ahead of the game.


Paul McGowan – PS Audio, Intl.

She turns into a tree?

Now that we have a rough understanding of the differences between classic digital audio (PCM) and the format Sony popularized in their SACD version of digital audio (DSD), it probably makes sense to explain why any of this even matters.

We  recognize that at the end of the day, all most of us want is to be able to sit and enjoy music on our stereo systems.  How that music is captured, stored and played back is of interest only when you’re digging deep to figure out what SHOULD work best for you and how to set an acceptable level of performance.  If you’re digging a track of music do you really care how it was recorded?

In my case, for example, the vast majority of my listening is on good old PCM CD’s, a few albums and the occasional high resolution digital.  Many of my favorite recordings and music are chosen not for their format – but how they sound and how they make me feel.  Much of that has to do with the music itself.  Occasionally I’ll find a recording of music that’s extremely well done but I simply don’t engage with the music.  In other cases I appreciate the music but the recording is so bad it just doesn’t work for me.  Do you feel the same?

My friend Jim McCullough of Cello recommended a track off of the Harmonia Mundi recording (Peter McGrath engineer) of Apollo and Dafne that I enjoy (track 8), but I listen to track 9 because it’s one of the loveliest pieces of music I know of, plus it is a tough one to playback properly.  Written by Handel, it features soprano Judith Nelson singing in harmony with the oboe.  The majesty of the piece depends entirely on the interplay between the two “instruments” working together.  When everything’s right, it’s easy to imagine how the composer must have wanted the performance to sound.  When not right, the singer and the instrument are disjointed and separate from each other.  One is magic, the other just good.

This is a standard PCM CD at 44.1kHz, 16 bit and it’s amazing.  Could it be better in DSD?  Better in a higher resolution?  Indeed, but I don’t want that to ever get in my way.

So, why does it matter to us the format, the technology, or any of that when a good old CD sounds just fine?

We care because the quest for better is built into most of us; certainly me.  I am not willing to give up something as lovely as the track I just described for technology, but if I can have better I will always go for it to the best of my ability.

That’s why it’s important to understand the boundary conditions we set for acceptability of what we care about – whether music, performance or both.


Paul McGowan – PS Audio, Intl.

Wait for it

Converting analog to DSD is a pretty simple process in some respects.  Even simpler than PCM.  But here’s a case where the concept may be simple but the execution is not.  However, in order to understand how it works, we don’t really need to wade into the details.

Basically we have a fixed clock running at 2.6MHz (single DSD, analogous to 176.4kHz in PCM) or twice that at 5.2MHz (double DSD, analogous to 352.8kHz in PCM).  This clock is just plugging along making its little square waves that will eventually become the actual bits that form the music.  In fact, the 1-bit we talk about is from this clock, producing a steady stream of 1-bits all at the same speed and interval.

So, picture a a guy standing at a gate.  Behind this gate is the steady stream of bits moving along at 2 million, 600 thousand times a second.  The gatekeeper’s job is to open the gate and let some of the bits through the gate whenever he hears some music.  If he hears no music, the gate is shut, none of the little bits can escape.  If he hears some music faintly playing in the background, he very carefully opens the gate and lets a few of the bits out.  The louder the music he hears, the more bits he lets out.

Now, add to this picture a neighbor to the gatekeeper that’s in on this scheme to release bits.  The neighbor can hear the music with one ear and with the other ear, he’s able to sense the number of bits being released by the gatekeeper.  His job is to compare the number of bits released with the loudness of the music and if there’s not enough, or if there’s too many, he signals the gatekeeper to release more or lessen the flow.

Between the gatekeeper releasing bits and his neighbor, keeping a watchful eye on his activities, the number of bits goes up and down in direct proportion to the music (the neighbor providing a constant feedback to keep the system correct).  That’s DSD or Pulse Density Modulation.  Now, remember a couple of things: the bits are always the same size, always the same speed.  All we’re doing is letting more (higher density) or less (lower density) through the gate.  If you just open the gate and let all the bits flow without any change, that’s as loud as the music can ever get.  Close the gate and let nothing through, that’s as quiet as it can get.

Keeping that picture of changing density in your mind and remembering that the bits are actually just energy – electrical energy – then it isn’t too big a leap to imagine that moving electrical energy can make music.  After all, isn’t analog just moving electrical energy?  Sure it is.  In fact, the only difference between the DSD moving energy and the analog moving energy is all those little stops and starts of the single bits.  How do you remove those and fill in the small gap between bits?

A DSD DAC which, in its simplest form, is nothing more than a resistor and a capacitor.  Remember?  Ten cents worth of parts go into a simple DSD DAC and those parts smooth out all the transistions and you are left with …… wait for it …….

Analog.


Paul McGowan – PS Audio, Intl.